Grounded in AMD CEO Lisa Su’s Complementarity Framework (February 2026)
Executive Summary
In February 2026, AMD Chief Executive Officer Lisa Su articulated a position that directly challenges the dominant displacement narrative in artificial intelligence discourse: that AI will not replace human labour but will augment it, making workers and organisations more productive. This case study examines that thesis in the context of Singapore — a small, open, knowledge-intensive economy that has been at the forefront of AI adoption in Asia — and analyses the outlook, policy solutions, and socioeconomic impacts of the augmentation paradigm for Singapore’s workforce and institutions.
“The idea that AI is going to replace everything, I am not a believer in that. I am a believer in: AI is going to make each of us better and each of our businesses better.” — Lisa Su, CEO, AMD, February 24, 2026
- Conceptual Framework: Augmentation vs. Substitution
1.1 The Substitution Thesis
The substitution model holds that AI systems will replicate cognitive tasks currently performed by human workers, rendering a significant proportion of white-collar employment redundant. This view gained traction following the emergence of large language models and generalist AI agents, and was amplified by market-moving reports (e.g., Citrini Research, 2026) positing that AI-driven displacement could destabilise equity markets and labour demand across a wide range of industries.
1.2 The Complementarity Model
Su’s position operationalises a complementarity model, which asserts that AI functions as a productivity-augmenting input rather than a direct substitute for labour. This is consistent with the task-based framework advanced by Acemoglu and Restrepo (2018, 2019), wherein automation displaces labour from specific tasks but simultaneously creates new tasks, reinstating demand for human labour at a higher level of cognitive complexity.
Three analytical claims underpin Su’s framework:
Human creativity and invention remain necessary prerequisite inputs in AI-mediated production processes.
AI accelerates output quality and throughput, creating efficiency rents that can be distributed across workers and firms.
Labour market adaptation — specifically reskilling toward ‘AI-forward’ competencies — is a necessary but manageable transition cost. - Singapore Context: The Augmentation Opportunity
2.1 Structural Position
Singapore occupies a structurally advantageous position relative to the augmentation paradigm. With a GDP per capita exceeding USD 87,000 (2024), near-full employment, and an economy concentrated in high-value services — finance, biomedical sciences, logistics, legal and professional services, and advanced manufacturing — Singapore’s labour force is disproportionately engaged in precisely those cognitive, non-routine tasks that stand to benefit most from AI augmentation.
The Singapore Smart Nation initiative, the National AI Strategy 2.0 (NAIS 2.0, 2023), and the Digital Economy Framework for Action collectively signal a policy orientation that is aligned with the augmentation thesis. Singapore has invested substantially in AI infrastructure, talent pipelines, and regulatory sandboxes, positioning itself as a regional AI hub.
2.2 Labour Market Profile
Sector AI Exposure Level Augmentation Potential Displacement Risk
Financial Services Very High High — data analytics, compliance, advisory Moderate (back-office)
Professional Services High High — research, drafting, due diligence Low-Moderate
Healthcare High High — diagnostics, clinical admin Low (human oversight req.)
Manufacturing (Advanced) High Very High — process optimisation, QC Moderate
Retail & Hospitality Moderate Moderate — customer service, inventory Moderate-High
Construction Low-Moderate Low — physical, site-variable tasks Low
Table 1: AI Exposure and Augmentation Potential by Sector in Singapore (Author’s Assessment, 2026)
2.3 The ‘AI-Forward’ Hiring Signal
Su’s disclosure that AMD has shifted hiring practices toward ‘AI-forward’ candidates — those who actively incorporate AI tools into their workflows — has direct relevance for Singapore’s human capital development strategy. The Ministry of Education (MOE), SkillsFuture Singapore (SSG), and the Infocomm Media Development Authority (IMDA) have all recognised this trend, with SkillsFuture Credit top-up schemes and AI literacy programmes increasingly targeting working adults across all education strata.
- Outlook: Scenarios for Singapore
3.1 Optimistic Scenario: Augmentation Dividend
Under the augmentation paradigm, Singapore’s relatively small but highly educated workforce achieves productivity gains commensurate with AI adoption. Firms in finance, law, and biomedical research leverage AI to expand output per employee, raising wages and sustaining employment. The government captures a share of this productivity surplus through progressive income taxation and the Skills Development Levy, redistributing resources toward continuous education.
In this scenario, the labour market bifurcates positively — demand rises for high-skill ‘AI orchestrators’ (those who direct and govern AI systems) while routine task workers successfully retrain through SkillsFuture programmes. Singapore’s position as an AI governance standard-setter in ASEAN strengthens, attracting foreign direct investment in AI-intensive industries.
3.2 Moderate Scenario: Frictional Transition
The more probable near-term scenario involves significant frictional unemployment concentrated in specific demographic and occupational cohorts. Workers in their 40s and 50s in administrative, paralegal, financial analysis, and customer service roles face the steepest adjustment costs. The risk is not permanent technological unemployment but structural unemployment during the transition — a period that, without adequate support, could entrench inequality between cohorts who successfully retrain and those who do not.
This scenario is consistent with Su’s acknowledgment that workers ‘might have to retrain’ — implying transition costs are real, even if not catastrophic.
3.3 Pessimistic Scenario: Displacement Without Redistribution
In the absence of proactive redistribution, the augmentation dividend accrues disproportionately to capital-owners and highly skilled AI-proficient workers, while mid-skill workers face wage suppression or displacement. Singapore’s already-notable income inequality (Gini coefficient approximately 0.437 before taxes and transfers, 2023) could widen, particularly if AI adoption in financial and professional services compresses mid-tier employment without commensurate wage gains for those who remain employed. - Policy Solutions: A Singapore-Specific Agenda
4.1 Human Capital and Reskilling
The SkillsFuture system provides a strong institutional foundation but requires recalibration for AI-era competencies. Specific recommendations include:
Expand AI literacy as a universal competency across all SkillsFuture-funded programmes, not merely as a specialist technical track.
Introduce ‘AI Augmentation Credits’ — targeted subsidies for mid-career professionals (40–55 years) in high-exposure occupations to access intensive AI workflow retraining.
Partner with firms like AMD, Google, and Salesforce to co-develop industry-validated AI competency frameworks aligned with ‘AI-forward’ hiring criteria.
Integrate practical AI tool proficiency (e.g., prompt engineering, agentic workflow management, AI output evaluation) into polytechnic and ITE curricula.
4.2 Labour Market Architecture
Singapore’s Progressive Wage Model (PWM) should be extended to incorporate an ‘AI Productivity Clause’ — a mechanism by which productivity gains attributable to AI augmentation in covered sectors are partially shared with workers through structured wage increments, preventing pure rent appropriation by capital owners.
The Fair Consideration Framework should be updated to require that employers demonstrate that AI-mediated hiring tools do not systematically disadvantage older workers or workers from lower socioeconomic backgrounds — an AI Fairness in Hiring standard analogous to existing anti-discrimination provisions.
4.3 Regulatory and Governance Frameworks
Singapore’s Model AI Governance Framework (2nd edition, 2020) should be revised to include sector-specific guidance on human-AI task allocation — establishing clear standards for which decisions require human oversight and which may be delegated to AI systems. This would operationalise Su’s principle that humans must ‘create and invent’ while AI executes and optimises.
Given AMD’s and other semiconductor companies’ pivotal role in AI infrastructure, Singapore should leverage its existing semiconductor ecosystem (GlobalFoundries, Micron, AMAT) to ensure domestic access to AI compute capacity — a strategic supply-chain resilience measure given ongoing geopolitical risk in chip supply chains.
4.4 Social Protection and Safety Nets
The Workforce Income Insurance Scheme (WIIS), currently voluntary, should be made mandatory for workers in high-AI-exposure occupations, funded through a co-payment structure between employers, employees, and the government. This provides income replacement during retraining transitions and reduces the political resistance to AI adoption that stems from individual economic insecurity. - Impact Assessment
5.1 Economic Impact
McKinsey Global Institute (2023) estimates that generative AI could add USD 2.6–4.4 trillion annually to global productivity. For Singapore, a conservative extrapolation suggests AI augmentation could contribute 1.5–2.5 percentage points to annual GDP growth over 2026–2035, contingent on effective adoption across high-value services. This is consistent with Singapore’s own projections under NAIS 2.0, which targets AI contributing S$1 billion in productivity gains by 2030.
5.2 Labour Market Impact
Impact Dimension Short-Term (2026–2028) Medium-Term (2029–2032)
Employment Level Stable; frictional displacement in select roles Net positive; new AI-adjacent roles created
Wage Distribution Modest wage growth for AI-proficient workers Wider premium; risk of polarisation if redistribution weak
Skills Demand Surge in AI literacy and prompt engineering demand Shift toward AI governance, creativity, judgment roles
Sectoral Composition Finance and professional services lead adoption Healthcare and manufacturing follow; all sectors impacted
Workforce Participation Slight dip due to frictional unemployment Recovery; older worker retention critical
Table 2: Projected Labour Market Impact in Singapore (Author’s Assessment, 2026)
5.3 Societal and Inequality Impact
The augmentation paradigm’s societal impact is contingent on distributional choices. Singapore’s institutional capacity — a sophisticated social compact built around meritocracy, continuous education, and strong tripartite labour relations (government–employers–unions) — positions it better than most economies to manage the transition. However, the historical tendency for technology-driven productivity gains to accrue disproportionately to capital and high-skill labour means proactive redistribution is not optional but essential.
Su’s framing — ‘we have to figure out how to harness the power of the technology instead of being afraid’ — resonates with Singapore’s national ethos of pragmatic adaptation. The City-State has previously navigated transformative economic shifts (industrialisation in the 1960s–70s, the services transition in the 1980s–90s, and the digital economy shift in the 2000s–10s) through proactive state-led investment in human capital and infrastructure. The AI transition represents the fourth such inflection point.
- Conclusion
Lisa Su’s complementarity thesis — that AI augments rather than replaces human labour — is not merely an optimistic assertion but a framework with substantial empirical and theoretical support, particularly when applied to a high-skill, knowledge-intensive economy such as Singapore. The key insight is that the outcome is not predetermined: whether AI produces an augmentation dividend or an inequality-amplifying displacement is a function of institutional design, policy choices, and the speed and inclusivity of workforce adaptation.
For Singapore, the augmentation paradigm represents both an economic opportunity of the first order and a governance challenge that will test the institutional resilience of its tripartite social compact. The solutions are known — expanded reskilling, progressive wage sharing, AI governance frameworks, and strong social safety nets during transition. The imperative is implementation at scale and speed commensurate with the pace of technological change.
In Su’s formulation, the charge to Singapore’s policymakers, educators, and business leaders is clear: do not fear the technology, but ensure that the structures exist to distribute its benefits equitably and to equip every worker to harness it.
References
Acemoglu, D., & Restrepo, P. (2018). The Race Between Man and Machine: Implications of Technology for Growth, Factor Shares, and Employment. American Economic Review, 108(6), 1488–1542.
Acemoglu, D., & Restrepo, P. (2019). Automation and New Tasks: How Technology Displaces and Reinstates Labor. Journal of Economic Perspectives, 33(2), 3–30.
Autor, D. H. (2015). Why Are There Still So Many Jobs? The History and Future of Workplace Automation. Journal of Economic Perspectives, 29(3), 3–30.
Greenberg, K. (2026, February 24). AMD Stock Jumped Today. CEO Lisa Su Doesn’t Think AI Will ‘Replace Everything’. Investopedia.
Infocomm Media Development Authority (IMDA). (2023). National AI Strategy 2.0. Singapore Government.
McKinsey Global Institute. (2023). The Economic Potential of Generative AI: The Next Productivity Frontier. McKinsey & Company.
Ministry of Finance, Singapore. (2023). Key Household Income Trends 2023. Department of Statistics Singapore.
Smart Nation and Digital Government Office (SNDGO). (2021). Digital Economy Framework for Action. Singapore Government.
SkillsFuture Singapore. (2024). SkillsFuture Mid-Career Enhanced Subsidy: Programme Overview. Singapore Government.